Journal of
Electrical and Electronics Engineering Research

  • Abbreviation: J. Electrical Electron. Eng. Res.
  • Language: English
  • ISSN: 2141-2367
  • DOI: 10.5897/JEEER
  • Start Year: 2009
  • Published Articles: 58

Full Length Research Paper

Efficacious approach for satellite image classification

Manish Sharma1*, Rashmi Gupta2, Deepak Kumar1 and Rajiv Kapoor2
  1Electronics and Communication Engineering Department, Maharaja Agrasen Institute of Technology, Sector -22, Rohini, Delhi -110086, India. 2Delhi Technological University (Formerly Delhi College of Engineering) Bawana Road, Delhi 110042, India.
Email: [email protected]

  •  Accepted: 28 February 2011
  •  Published: 31 October 2011

Abstract

 

The main idea behind any image classification process is to obtain highest accuracy possible. Minimum distance and parallelepiped method yielded acceptable results for image classification but they are bounded by their inherent limitations. On the other hand, fuzzy based systems are fast and provide good accuracy. In fuzzy, accuracy depends upon the type of membership function used, and how the membership functions in the output of FIS are arranged. In this paper Mamdani fuzzy inference system is used to classify image and how the arrangements and the type of fuzzy membership functions employed in the classification, affected the results obtained, are shown.

 

Key words: Image classification, fuzzy logic, type of membership function, positioning of membership functions